signal-bot/signal_bot/stats/render_extra.py
Krzysztof kuhy Rudnicki 40538b89a1 fix: !ja told everybody their favourite hour was noon
Reported from the group: the same "Najczęściej piszesz między 12:00 a
13:00" line came back for Ala and for Jan Nowak, which is too much
of a coincidence to be a statistic.

It was the backfill. The 342 messages carried over from the pre-0.1.0
counter have no timestamps, so the seeded rows got a placeholder hour of
12 -- and favourite_hour was the one query that never excluded the
backfilled day. For every person those synthetic noons outnumbered every
real message they had sent since, so all seven got the same answer. The
genuine hours are 19-21: this group talks in the evening.

The comment above that placeholder said "the day is marked anyway, so
nothing that cares about hours will read these", which is exactly the
kind of reasoning that stops being true the moment somebody adds a
query. So the fix is not another `exclude` argument to remember: the
absent hour is now -1, which no real hour can be, and hour queries
filter `local_hour >= 0`. Forgetting the exclusion can no longer invent
an answer, and night_owl gets the same guard even though its 0..2 window
happened to miss the old placeholder.

Live database migrated with the service stopped and a backup taken: 342
rows moved from noon to unknown. Verified afterwards on the real data --
Ala 19:00, Jan Nowak 19:00, kuhy 20:00, and the two people whose
only messages are backfilled correctly get no hour line at all.

Not released. Sitting at 0.1.1 until kuhy says otherwise.

Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015d48t4qsQLBJnDwDeo2Nom
2026-08-30 20:15:20 +02:00

195 lines
7.3 KiB
Python

# Copyright (c) 2026 Krzysztof Rudnicki. MIT licence.
"""The on-demand answers: records, the crown, one person, the weekdays.
Each of these is asked for rather than pushed, so each says something even
when there is nothing interesting yet -- "za mało danych" is a useful answer
to a question, where it would be noise in a nightly post nobody asked for.
"""
from __future__ import annotations
from typing import TYPE_CHECKING, Final
from signal_bot.stats import records, streaks, text
from signal_bot.stats.periods import WEEKDAY_PLURAL, pl_day_month
if TYPE_CHECKING:
from signal_bot.stats.events import EventStore, Person
from signal_bot.stats.records import DayRecord
TROPHY: Final = "\U0001f3c6"
CROWN: Final = "\U0001f451"
CALENDAR: Final = "\U0001f4c5"
CAMERA: Final = "\U0001f4f7"
QUIET: Final = "\U0001f92b"
HOURS_IN_A_DAY: Final = 24
# Every command says what it is, whether or not it has data yet. Somebody
# typing !korona on day one used to get "za mało danych" and nothing else,
# which tells them neither what the command does nor when it will work.
TITLES: Final = {
"crown": f"{CROWN} Korona — kto danego dnia napisał najwięcej wiadomości",
"records": f"{TROPHY} Rekordy — najlepsze i najgorsze dni w historii grupy",
"weekdays": f"{CALENDAR} Dni tygodnia — ile średnio piszecie w każdy z nich",
}
# What each will show once there is enough history, so the empty answer is
# still an answer.
PROMISES: Final = {
"crown": (
"Pokażę: kto ma koronę teraz, jak długą serię, kto miał najdłuższą "
"w historii i ile dni każdy ją nosił."
),
"records": (
"Pokażę: najgadatliwszy dzień, najcichszy, dzień z największą liczbą "
"zdjęć, najdłuższą ciszę i rekord życiowy każdego z was."
),
"weekdays": (
"Pokażę: średnią liczbę wiadomości w poniedziałki, wtorki i tak dalej "
"— czyli czy weekendy są gadatliwsze od środy."
),
}
def nothing_yet(what: str, since: str | None = None) -> str:
"""The empty answer, which still says what the command is for.
A bare "za mało danych" tells somebody neither what they asked for nor
when it will work, which is exactly what happened within minutes of
0.1.0 going out.
"""
started = f"Liczę od {pl_day_month(since)}. " if since else ""
return "\n".join([TITLES[what], "", f"{started}Jeszcze za mało danych.", PROMISES[what]])
def _record_line(label: str, record: DayRecord | None, unit: str) -> str | None:
"""One "label: value (date)" line, or None when there is no record yet."""
if record is None:
return None
return f"{label}: {record.value} {unit} ({pl_day_month(record.day)})"
def render_records(
history: dict[str, int],
media: DayRecord | None,
silence: records.Silence | None,
personal: list[tuple[str, DayRecord]],
since: str | None = None,
) -> str:
"""Every all-time record the group has set."""
lines = [TITLES["records"]]
candidates = [
_record_line("Najwięcej w jeden dzień", records.busiest_day(history), "wiadomości"),
_record_line("Najciszej (ale ktoś pisał)", records.quietest_day(history), "wiadomości"),
_record_line(f"Najwięcej zdjęć {CAMERA}", media, "zdjęć"),
]
lines.extend(line for line in candidates if line)
if silence:
lines.append(
f"{QUIET} Najdłuższa cisza: {text.days(silence.days)} "
f"({pl_day_month(silence.start)} - {pl_day_month(silence.end)})"
)
if personal:
lines.append("")
lines.append("Rekordy życiowe:")
lines.extend(
f" {name} — {record.value} ({pl_day_month(record.day)})" for name, record in personal
)
return "\n".join(lines) if len(lines) > 1 else nothing_yet("records", since)
def render_crown(
winners: dict[str, list[tuple[str, str]]],
people: list[Person],
since: str | None = None,
) -> str:
"""Who holds the crown now, who held it longest, and who has held it most."""
if not winners:
return nothing_yet("crown", since)
# Both are guaranteed once winners is non-empty, which the guard above
# already established: the latest day always has at least one holder, and
# a day with a holder is a run of at least one.
current = streaks.current_streaks(winners)
# `or current[0]` is a type narrowing, not a fallback: longest_ever only
# returns None for an empty history, which the guard above ruled out.
longest = streaks.longest_ever(winners) or current[0]
names = " i ".join(s.display_name for s in current if s.length == current[0].length)
lines = [
TITLES["crown"],
f"Teraz: {names} — {text.days(current[0].length)} z rzędu",
(
f"Najdłuższa seria: {longest.display_name} — {text.days(longest.length)} "
f"(do {pl_day_month(longest.last_day)})"
),
]
held = sorted(
((person.display_name, streaks.held_days(winners, person.uuid)) for person in people),
key=lambda pair: (-pair[1], pair[0]),
)
lines.append("")
lines.append("Dni z koroną łącznie:")
lines.extend(f" {name} — {text.days(days)}" for name, days in held if days)
return "\n".join(lines)
def render_weekdays(buckets: dict[int, list[int]], today: int, since: str | None = None) -> str:
"""The Monday-to-Sunday picture, with today's weekday called out."""
if not buckets:
return nothing_yet("weekdays", since)
lines = [TITLES["weekdays"]]
for index in range(len(WEEKDAY_PLURAL)):
days = buckets.get(index)
if not days:
continue
mean = round(sum(days) / len(days))
marker = " ← dziś" if index == today else ""
lines.append(
f"{WEEKDAY_PLURAL[index]}: średnio {mean} ({text.days(len(days))} w historii){marker}"
)
return "\n".join(lines)
ME_TITLE: Final = "Twoje liczby w tej grupie"
def render_me(
name: str,
total: int,
best: DayRecord | None,
share: int,
favourite_hour: int | None,
streak: int,
) -> str:
"""One person's own numbers."""
lines = [ME_TITLE, "", f"{name} — {text.messages(total)} w historii"]
if best:
lines.append(f"{TROPHY} Rekord życiowy: {best.value} ({pl_day_month(best.day)})")
lines.append(f"Udział w rozmowie: {share}%")
if favourite_hour is not None:
lines.append(
f"Najczęściej piszesz między {favourite_hour:02d}:00 a "
f"{(favourite_hour + 1) % HOURS_IN_A_DAY:02d}:00"
)
if streak:
lines.append(f"{CROWN} Korona: {text.days(streak)} z rzędu")
return "\n".join(lines)
def favourite_hour(events: EventStore, group_id: str, uuid: str) -> int | None:
"""The hour somebody writes most in, or None when nothing is known.
Skips rows whose hour is UNKNOWN_HOUR. Those are backfilled from a counter
that had no timestamps, and counting them told every member of the group
that they write most at noon.
"""
row = events.connection.execute(
"""
SELECT local_hour, COUNT(*) AS messages
FROM message_events
WHERE group_id = ? AND uuid = ? AND local_hour >= 0
GROUP BY local_hour
ORDER BY messages DESC, local_hour ASC
LIMIT 1
""",
(group_id, uuid),
).fetchone()
return row["local_hour"] if row else None